1. Field of the Invention
This invention relates to a waveform generating method, a performance data processing method, a waveform selection apparatus, a waveform data recording apparatus, and a waveform data recording and reproducing apparatus for use in making performance using recorded musical instrument tones and singing tones.
2. Description of the Related Art
Conventionally, there is known a waveform memory tone generator. The waveform memory tone generator generates musical tones by reading out waveform data stored in a waveform memory according to event data such as note-on and note-off. In reading waveform data from the waveform memory, different waveform data are selected from the waveform memory according to pitch information and touch information contained in note-on events.
On the other hand, an electronic musical instrument called a sampler is also conventionally known. The sampler samples and records waveforms of monotones being performed (for example, a pitch C3 for five seconds) and generates musical tones using the recorded waveform data. In this case, the user sets original keys representing tone pitches of respective ones of the sampled waveform data, for the respective sampled waveform data. A range where the waveform data is to be used, an amount of pitch shift during sounding at a predetermined pitch, and the like are determined according to the original keys. To determine the original keys, pitches are sampled from the sampled waveform data and the original keys are automatically set for the waveform data correspondingly to the sampled pitches (refer to Japanese Laid-Open Patent Publication (Kokai) No. 07-325579, for example).
A phrase sampler has been also proposed which samples phrase waveforms being performed, divides each of the sampled phrase waveforms into a plurality of partial waveform data according to an envelop level sampled from the sampled phrase waveform data to make performance while changing timing and pitch of each partial waveform data.
If tones being actually generated by performance are sampled and recorded and musical tones are generated using the recorded waveform data as is the case with the conventional sampler, tones are generated with poorer expression than tones actually generated by performance by musical instruments. In this case, it is impossible to obtain real musical tones even if waveform data is selected according to pitch information and touch information.
Further, according to the conventional sampler, since each waveform data for use in performance is recorded in the form of monotones, tones generated based upon the recorded waveform data are unnatural as is different from tones produced during performance of music. Specifically, the player needs to sequentially perform and record necessary monotones in a sampler during recording, but the player cannot easily perform only a monotone without getting nervous as compared with performance of a phrase. Particularly when a single tone is sung as a vocal, a voice may be uttered poorly or in a falsetto tone. Since it is difficult to record monotones in natural tone color, natural tones cannot be easily generated by performance by the sampler.
Further, when setting original keys for the recorded monotones, a person must determine the original keys or pitches must be sampled to determine the original keys. He or she, however, must be experienced to determine the original keys, and therefore, every person cannot determine the original keys. On the other hand, sampling the pitches requires complicated arithmetic operations, and moreover, the sampled pitches may cause one tone being generated by performance to be incorrectly recognized as multiple tones due to a change in pitch over time or may cause any of harmonic tones existing in some tones rich with harmonic components being generated by performance to be incorrectly determined as fundamental tones. Thus, the sampled pitches cannot always be correct. That is, even if the pitches are sampled, a person must finally confirm the sampled pitches.
Further, the conventional phrase sampler does not sample a pitch from waveform data when dividing a phrase waveform into partial waveforms. More specifically, a position at which a phrase waveform is divided into partial waveforms does not necessarily correspond to a point in change of pitch, and it is therefore impossible to set original keys for respective ones of partial waveform data. Further, to set original keys for partial waveform data with the dividing position being regarded as a point of change in pitch, the user needs to set the original keys or sample pitches from waveform data as mentioned above. This requires complicated operations as described above.
It is therefore a first object of the present invention to provide a waveform generating method, a performance data processing method, a waveform selection apparatus, a waveform data recording apparatus, and a recorded waveform reproducing apparatus, which are capable of generating expressive musical tones.
It is a second object of the present invention to provide a waveform generating method, a performance data processing method, a waveform selection apparatus, a waveform data recording apparatus, and a recorded waveform reproducing apparatus, which make it possible to record natural tones, divide recorded waveform data at points of change in pitch, and automatically assigning property information on pitch and the like to the divided waveform data.
It is a third object of the present invention to provide a waveform generating method, a performance data processing method, a waveform selection apparatus, a waveform data recording apparatus, and a recorded waveform reproducing apparatus, which are capable of obtaining natural performance tones.
To attain the first object, the present invention provides a waveform generating method comprising the steps of storing a plurality of partial waveforms in a partial waveform memory, storing property information on respective ones of the partial waveforms stored in the partial waveform memory, in a property information memory, retrieving the property information memory according to inputted sounding control information to read out a partial waveform having property information corresponding to the sounding control information, processing the readout partial waveform according to the property information and the sounding control information, and generating a waveform corresponding to the sounding control information.
To attain the first object, the present invention also provides a performance data processing method comprising the steps of storing property information on respective ones of a plurality of partial waveforms in a property information memory, comparing characteristics of respective ones of notes included in performance data with the property information stored in the property information memory to detect an optimum partial waveform for a characteristic of each of the notes, assigning designation data for designating the detected partial waveform to each of the notes, and storing performance data having the designation data assigned thereto.
To attain the first object, the present invention further provides a waveform generating method for use in reproducing the performance data having the designation data assigned thereto stored by the above-mentioned performance data processing method, comprising the steps of storing a plurality of partial waveforms in a partial waveform memory, reading out a partial waveform from the partial waveform memory according to the designation data assigned to each of the notes to be reproduced when reproducing the notes having the designation data assigned thereto according to the performance data, and generating a waveform corresponding to each of the notes based upon the read out partial waveform.
To attain the first object, the present invention yet further provides a waveform generating method comprising the steps of storing performance information corresponding to real-time performance, and generating accompanying tones by automatic performance or automatic accompaniment according to a tempo clock, reproducing performance information according to the tempo clock, and generating waveforms corresponding to performance events performed in real time to accompaniment of the generated accompanying tones, according to the performance events and the reproduced performance information.
To attain the first object, the present invention further provides a waveform selection apparatus comprising a partial waveform memory that stores a plurality of partial waveforms, a database that stores property data representing characteristics of two tones corresponding to each of the partial waveforms stored in the partial waveform memory and consisting of a tone corresponding to each of the partial waveforms and a preceding tone, and a retrieving device that retrieves the database according to characteristic data of two tones consisting of an inputted present tone to be generated by performance and an inputted preceding tone to be generated by performance before the inputted present tone to extract at least one partial waveform having property data close to the characteristic data of the inputted two tones as a partial waveform for sounding the inputted present tone to be generated by performance.
To attain the first object, the present invention further provides a waveform selection apparatus comprising a partial waveform memory that stores a plurality of partial waveforms, a database that stores property data representing characteristics of two tones corresponding to each of the partial waveforms stored in the partial waveform memory and consisting of a tone corresponding to each of the partial waveforms and a following tone, and a retrieving device that retrieves the database according to characteristic data of two tones consisting of an inputted present tone to be generated by performance and a following tone to be inputted and generated by performance before the inputted present tone to extract at least one partial waveform having property data close to the characteristic data of the inputted present tone and the following tone to be inputted as a partial waveform for sounding the inputted present tone to be generated by performance.
According to the waveform generating method of the present invention, it is possible to generate waveforms using the optimum partial waveforms correspondingly to sounding control information, and hence generate expressive musical tones.
Further, according to the performance data processing method of the present invention, it is possible to enable the optimum partial waveform data for characteristics of each note in the performance data to be selected in advance and assigned to each note in the performance data. The use of such performance data eliminates the necessity of selecting the optimum partial waveform data during waveform generation.
Further, according to the waveform generating method of the present invention, even when musical tones are generated by real-time performance, it is possible to control characteristics of musical tones generated by performance being currently made according to following performance to be subsequently made, by utilizing performance information corresponding to real-time performance.
Further, by selecting one or more partial waveforms having property data close to characteristic data of the following two tones, a present tone to be currently generated by performance and a preceding tone to be generated by performance before the present tone or a following tone to be generated by performance next to the present tone, the optimum partial waveforms can be selected as partial waveforms for generating musical tones by performance.
To attain the second object, the present invention provides a waveform data recording apparatus comprising an automatic performing device that reproduces performance data relating to phrases to be recorded, a waveform recording device that records phrase waveforms representing tones generated by performance based on tones generated by reproduction of the performance data by the automatic performing device, and a waveform data processing device that extracts data of the phrase waveforms recorded by the waveform recording device according to characteristic information on notes of the performance data to divide the data of the phrase waveforms into partial waveform data corresponding to respective ones of the notes.
Preferably, the waveform recording device records the phrase waveforms in synchronism with performance timing of the automatic performing device.
Also preferably, the waveform data processing device assigns the characteristic information on the notes corresponding to the partial waveform data as property information to the partial waveform data.
Preferably, the waveform data recording apparatus comprises a device that creates a tone color set from a partial waveform pool in which selected partial waveform data selected from the partial waveform data is pooled, and a partial waveform management database composed of the property information assigned to respective ones of the selected partial waveform data pooled in the partial waveform pool.
To attain the third object, the present invention also provides a recorded waveform data reproducing apparatus comprising a storage device that stores a tone color set from a partial waveform pool in which selected partial waveform data selected from the partial waveform data is pooled, and a partial waveform management database composed of property information assigned to respective ones of the selected partial waveform data pooled in the partial waveform pool, a detecting device that retrieves the partial waveform management database in the tone color set according to characteristic information on respective notes in performance data to detect optimum partial waveform data for each of the notes from the partial waveform pool, a designation data inserting device that embeds designation data for designating the optimum partial waveform data detected by the detecting device to each of the notes into the performance data, and a reproducing device that automatically reproduces the performance data in which the designation data is embedded by the designation data inserting device, according to the optimum partial waveform data designated by the designation data.
To attain the third object, the present invention further provides a recorded waveform data reproducing apparatus comprising a storage device that stores a tone color set from a partial waveform pool in which selected partial waveform data selected from the partial waveform data is pooled, and a partial waveform management database composed of property information assigned to respective ones of the selected partial waveform data pooled in the partial waveform pool, a detecting device that retrieves the partial waveform management database in the tone color set according to information on respective performance event data that have occurred, to detect optimum partial waveform data for each of the performance event data from the partial waveform pool, and a reproducing device that reproduces the performance event data according to the optimum partial waveform data detected by the detecting device.
To attain the third object, the present invention yet further provides a recorded waveform data reproducing apparatus comprising a storage device that stores a tone color set from a partial waveform pool in which selected partial waveform data selected from the partial waveform data is pooled, and a partial waveform management database composed of property information assigned to respective ones of the selected partial waveform data pooled in the partial waveform pool, a detecting device that retrieves the partial waveform management database in the tone color set according to characteristic information on respective notes in performance data read in advance to detect optimum partial waveform data for each of the notes from the partial waveform pool, and a reproducing device that is responsive to occurrence of performance event data corresponding to respective ones of the notes in the performance data, for reproducing performance tones corresponding to the respective ones of the notes according to the optimum partial waveform data detected by the detecting device.
According to the present invention constructed as above, it is possible to enable performance to be made while listening to tones generated by performance by the automatic performing device, and thus enable the player to make performance relaxedly and make recording of natural tones. By extracting the thus recorded phrase waveform data according to the automatically reproduced performance data, the phrase waveform data can be divided into partial waveform data corresponding to notes of the performance data. In this case, the dividing position is made more accurate by recording tones generated by performance in synchronism with the performance timing of the automatic performing device. Further, the characteristic information such as the pitch, length, intensity of a note corresponding to the partial waveform data obtained by the division can be assigned as property information to the corresponding partial waveform data. Further, desired partial waveform data can be selected from the partial waveform data obtained by the division and combined with the property information thereof to provide the tone color set for performance.
To make automatic performance based on the tone color set, the partial waveform management database of the tone color set is retrieved according to the pitch, length, intensity, etc. of each note in performance data for automatic performance to detect the optimum partial waveform data, and designation data for designating the detected partial waveform data to the note is embedded in the performance data. This enables automatic performance to be made with natural tones based on the processed performance data.
Further, to make real-time performance based on the tone color set, the partial waveform management data of the tone color set is retrieved according to information on generated performance event data to detect the optimum partial waveform, and the detected partial waveform data is used to reproduce performance tones of the performance event data, to thereby enable performance to be made with natural tones.
Further, to make performance based on the tone color set, the partial waveform management database of the tone color set is retrived according to the pitch, length, intensity, etc. of each note in performance data read in advance to detect the optimum partial waveform data, and upon occurrence of performance event data corresponding to respective ones of the notes in the performance data when the same part as the performance data is performed, performance tones corresponding to the respective notes are reproduced according to the detected optimum partial waveform data. This enables detection of conditions of following tones to be reproduced and hence enables performance to be made with more natural tones.
The above and other objects, features, and advantages of the invention will become more apparent from the following detailed description taken in conjunction with the accompanying drawings.